Plasma Engine  2.0
Loading...
Searching...
No Matches
plSeparatedBilateralBlurPass Class Reference

Depth aware blur on input and writes it to an output buffer of the same format. More...

#include <SeparatedBilateralBlur.h>

Inheritance diagram for plSeparatedBilateralBlurPass:

Public Member Functions

virtual bool GetRenderTargetDescriptions (const plView &view, const plArrayPtr< plGALTextureCreationDescription *const > inputs, plArrayPtr< plGALTextureCreationDescription > outputs) override
 For a given input pin configuration, provide the output configuration of this node. Outputs is already resized to the number of output pins.
 
virtual void Execute (const plRenderViewContext &renderViewContext, const plArrayPtr< plRenderPipelinePassConnection *const > inputs, const plArrayPtr< plRenderPipelinePassConnection *const > outputs) override
 Render into outputs. Both inputs and outputs are passed in with actual texture handles. Disconnected pins have a nullptr value in the passed in arrays. You can now create views and render target setups on the fly and fill the output targets with data.
 
virtual plResult Serialize (plStreamWriter &inout_stream) const override
 
virtual plResult Deserialize (plStreamReader &inout_stream) override
 
void SetRadius (plUInt32 uiRadius)
 
plUInt32 GetRadius () const
 
void SetGaussianSigma (float fSigma)
 
float GetGaussianSigma () const
 
void SetSharpness (float fSharpness)
 
float GetSharpness () const
 
- Public Member Functions inherited from plRenderPipelinePass
 plRenderPipelinePass (const char *szName, bool bIsStereoAware=false)
 
void SetName (const char *szName)
 Sets the name of the pass.
 
const char * GetName () const
 returns the name of the pass.
 
bool IsStereoAware () const
 True if the render pipeline pass can handle stereo cameras correctly.
 
virtual plGALTextureHandle QueryTextureProvider (const plRenderPipelineNodePin *pPin, const plGALTextureCreationDescription &desc)
 
virtual void InitRenderPipelinePass (const plArrayPtr< plRenderPipelinePassConnection *const > inputs, const plArrayPtr< plRenderPipelinePassConnection *const > outputs)
 After GetRenderTargetDescriptions was called successfully for each pass, this function is called with the inputs and outputs for review. Disconnected pins have a nullptr value in the passed in arrays. This is the time to create additional resources that are not covered by the pins automatically, e.g. a picking texture or eye adaptation buffer.
 
virtual void ExecuteInactive (const plRenderViewContext &renderViewContext, const plArrayPtr< plRenderPipelinePassConnection *const > inputs, const plArrayPtr< plRenderPipelinePassConnection *const > outputs)
 
virtual void ReadBackProperties (plView *pView)
 Allows for the pass to write data back using plView::SetRenderPassReadBackProperty. E.g. picking results etc.
 
void RenderDataWithCategory (const plRenderViewContext &renderViewContext, plRenderData::Category category, plRenderDataBatch::Filter filter=plRenderDataBatch::Filter())
 
PL_ALWAYS_INLINE plRenderPipelineGetPipeline ()
 
PL_ALWAYS_INLINE const plRenderPipelineGetPipeline () const
 
- Public Member Functions inherited from plRenderPipelineNode
void InitializePins ()
 
plHashedString GetPinName (const plRenderPipelineNodePin *pPin) const
 
const plRenderPipelineNodePinGetPinByName (const char *szName) const
 
const plRenderPipelineNodePinGetPinByName (plHashedString sName) const
 
const plArrayPtr< const plRenderPipelineNodePin *const > GetInputPins () const
 
const plArrayPtr< const plRenderPipelineNodePin *const > GetOutputPins () const
 
- Public Member Functions inherited from plReflectedClass
virtual const plRTTIGetDynamicRTTI () const
 
bool IsInstanceOf (const plRTTI *pType) const
 Returns whether the type of this instance is of the given type or derived from it.
 
template<typename T >
PL_ALWAYS_INLINE bool IsInstanceOf () const
 Returns whether the type of this instance is of the given type or derived from it.
 

Protected Attributes

plRenderPipelineNodeInputPin m_PinBlurSourceInput
 
plRenderPipelineNodeInputPin m_PinDepthInput
 
plRenderPipelineNodeOutputPin m_PinOutput
 
plUInt32 m_uiRadius = 7
 
float m_fGaussianSigma = 3.5f
 
float m_fSharpness = 120.0f
 
plConstantBufferStorageHandle m_hBilateralBlurCB
 
plShaderResourceHandle m_hShader
 

Additional Inherited Members

- Static Public Member Functions inherited from plNoBase
static const plRTTIGetStaticRTTI ()
 

Detailed Description

Depth aware blur on input and writes it to an output buffer of the same format.

In theory it is mathematical nonsense to separate a bilateral blur, but it is common praxis and works good enough. (Thus the name "separated" in contrast to "separable")

Member Function Documentation

◆ Deserialize()

plResult plSeparatedBilateralBlurPass::Deserialize ( plStreamReader & inout_stream)
overridevirtual

Reimplemented from plRenderPipelinePass.

◆ Execute()

void plSeparatedBilateralBlurPass::Execute ( const plRenderViewContext & renderViewContext,
const plArrayPtr< plRenderPipelinePassConnection *const > inputs,
const plArrayPtr< plRenderPipelinePassConnection *const > outputs )
overridevirtual

Render into outputs. Both inputs and outputs are passed in with actual texture handles. Disconnected pins have a nullptr value in the passed in arrays. You can now create views and render target setups on the fly and fill the output targets with data.

Implements plRenderPipelinePass.

◆ GetRenderTargetDescriptions()

bool plSeparatedBilateralBlurPass::GetRenderTargetDescriptions ( const plView & view,
const plArrayPtr< plGALTextureCreationDescription *const > inputs,
plArrayPtr< plGALTextureCreationDescription > outputs )
overridevirtual

For a given input pin configuration, provide the output configuration of this node. Outputs is already resized to the number of output pins.

Implements plRenderPipelinePass.

◆ Serialize()

plResult plSeparatedBilateralBlurPass::Serialize ( plStreamWriter & inout_stream) const
overridevirtual

Reimplemented from plRenderPipelinePass.


The documentation for this class was generated from the following files: